AI

Our research examines vulnerabilities that affect Large Language Model (LLM) powered agents with code execution, document upload, and internet access capabilities. This is the second part of a series diving into the critical vulnerabilities in AI agents.

This introductory post kicks off a blog series on AI agent vulnerabilities, outlining key security risks like prompt injection and code execution, and sets the stage for future parts, which will dive deeper into issues such as code execution flaws, data exfiltration, and database access threats.

The Trend 2025 Cyber Risk Report sustains our shift towards proactive security. Protecting enterprises is no longer about stopping breaches but is now about staying ahead, making cybersecurity a business enabler. By looking at the 2025 risk landscape, we recognize exposures and understand attacker behavior to be able to implement countermeasures, transforming security from a challenge to a catalyst for innovation and business growth.

This article offers a fresh perspective on AI security, emphasizing that while core AI models are often the focus, the surrounding infrastructure and cloud services are equally critical to securing AI applications, particularly with the rise of AI-as-a-Service.

Consumer data will be the hot commodity in the underground for 2025, with cybercrime expected to cost over $10 trillion in this coming year. Criminals will continue to develop new ways to exploit vulnerable areas, increasing enterprise risk as the attack surface expands.

Digital Assistants (DA) are AI-driven software, sometimes embedded into dedicated hardware and integrated with multiple systems, that understand natural language and use them to perform various tasks

How can misconfigurations help threat actors abuse AI to launch hard-to-detect attacks with massive impact? We reveal how AI models stored in exposed container registries could be tampered with— and how organizations can protect their systems.

This analysis investigates the security risks of eKYC systems in relation to deepfake attacks, highlighting the diverse strategies employed by cybercriminals in bypassing eKYC security measures.

Enterprises have been steadily adopting private blockchains, placing this technology in different areas of their operations. When deployed as private permissioned networks, Ethereum-based blockchains and other similar decentralized ledger technologies are beneficial for enterprises. The advantages of integrating private blockchains into organizational processes can be substantial and multifaceted.